Hadoop-based Distributed RSA Encryption Rapid Design And Implementation Of The Algorithm

Nowadays, distributed cloud platform has become an important developing tendency. The current main IT companies such as Baidu, Tencent, Google, Alibaba and IBM have individually constructed their "private cloud computing platform," and conduct wide open research based on the platforms. In this paper, we follow the trend of world IT industry, constructing a distributed platform using the lab server, and developing an individual cloud distributed platform. The system constructed is based on a Hadoop open source project of Apache. We implemented a discussion and research of "distributed RSA algorithm" based on the distributed file system of Hadoop (HDFS) and the distributed computation model (MapReduce).Beginning of this paper, two core distributed file systems called HDFS and distributed computation model MapReduce have been deeply researched and discussed. The principle of RSA encryption algorithm has been researched and a discussion on the drawbacks of the algorithm has been proposed. Based on the researches mentioned above, our research has been divided into two parts.In the part III of research, a widely used algorithm in RSA encryption algorithm, named large integer multiplication algorithm, has been discussed. Numerous large integer multiplication algorithms have been analyzed and an improved algorithm has been proposed.In the part III of research, this paper proposed a specific design procedure for distributed RSA algorithm based on Hadoop platform. The distributed encryption function is achieved through defining configuration of distributed RSA algorithm, modeling the algorithm and controlling the procedure. The data types and store method of plaintext, ciphertext, public key and private key of RSA encryption have been designed, and a specific implementation of each model in RSA coding has been completed. A functional test, expansibility test and efficiency test for the RSA encryption algorithm has been conducted through constructed massive distributed cluster in the lab. Based on the measurement.data, the load balance problem for Hadoop distributed cluster and an optimization problem for Slaver amount selection have been discussed. A future development and research toward this problem has also been discussed.